Diabetes mellitus is a common disorder with a prevalence of 4-5 percent and is classified into three major forms: type 1 diabetes often referred to as immune mediated diabetes, type 2 diabetes or non-insulin dependent diabetes, and diabetes due to mutations in genes controlling beta cell function or metabolism. The effect of age is marked and type 1 diabetes dominates among children and teenagers. The etiology and pathogenesis of type 1 diabetes is strongly associated with autoreactivity to glutamic acid decarboxylase 65 (GAD65), insulin or both. This research will test the hypothesis that constitutive expression of secretable GAD65 or proinsulin will tolerize diabetes prone rats and prevent or significantly delay the onset of diabetes. As a contingency plan, should single administration of autoantigen not be effective in inducing tolerance, we will co-express both GAD65 and proinsulin from bicistronic vectors. BB rats with the lyp/lyp genotype predictably develop diabetes between 60-90 days of age permitting the delivery of islet autoantigens before disease onset to study tolerance induction and protection from diabetes. Diabetes is prevented or delayed by injecting young diabetes-prone NOD mice with either one of these two autoantigens and we have shown T cell cytokine deviation in response to GAD65 injections in the BB rat. We believe that utilizing gene therapy for the constitutive expression of autoantigens presents a significant improvement over serial administration of autoantigen by injection. We propose to investigate 2 Specific Aims. In aim 1 we will construct novel retroviral vectors encoding secretable GAD65 and proinsulin under the control of a fibronectin promoter to transduce skin fibroblasts for long-term expression from cells introduced as skin equivalent grafts to control and BB diabetic rats. Regulation of autoantigen delivery will be achieved by controlling the number of fibroblasts implanted. When tolerance has been achieved skin grafts will be removed to monitor preservation of this altered immune state in the absence of autoantigen. In aim 2 we will generate HIV-1 based lentiviral vectors expressing proinsulin or secretable rat GAD65 elements for direct intramuscular injection into prediabetic BB rats. The levels of autoantigen delivered will be controlled by variation in number of lentivirus particles administered. The overall goal of this application is the development of a gene therapy method of tolerance induction that can be applied to patients with type 1 diabetes and their high risk relatives, and may be applied to other autoimmune diseases.